Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
[Preprint]. 2024 Dec 6:2024.12.04.24318520.
doi: 10.1101/2024.12.04.24318520.

Using genomic epidemiology and geographic activity spaces to investigate tuberculosis outbreaks in Botswana

Chelsea R Baker  1 , Ivan Barilar  2 , Leonardo S de Araujo  2 , Daniel M Parker  1 , Kimberly Fornace  3 , Patrick K Moonan  4 , John E Oeltmann  4 , James L Tobias  4 , Volodymyr M Minin  1 , Chawangwa Modongo  5 , Nicola M Zetola  5 , Stefan Niemann  2 , Sanghyuk S Shin  1
Affiliations

Using genomic epidemiology and geographic activity spaces to investigate tuberculosis outbreaks in Botswana

Chelsea R Baker et al. medRxiv. .

Abstract

Background: The integration of genomic and geospatial data into infectious disease transmission analyses typically includes residential locations and excludes other activity spaces where transmission may occur (e.g. work, school, or social venues). The objective of this analysis was to explore residential as well as other activity spaces of tuberculosis (TB) outbreaks to identify potential geospatial 'hotspots' of transmission.

Methods: We analyzed data that included geospatial coordinates for residence and other activity spaces collected during 2012-2016 for the Kopanyo Study, a population-based study of TB transmission in Botswana. We included participants with results from whole genome sequencing conducted on archived samples from the original study. We used a spatial log-Gaussian Cox process model to detect core areas of increased activity spaces of individuals belonging to TB outbreaks (genotypic groups with ≤5 single-nucleotide polymorphisms), which we compared to ungrouped participants (those not in a genotypic group of any size).

Findings: We analyzed data collected from 636 participants, including 70 participants belonging to six outbreak groups with a combined total of 293 locations, and 566 ungrouped participants with a combined total of 2289 locations. Core areas of activity space for each outbreak group were geographically distinct, and we found evidence of localized transmission in four of six outbreaks. For most of the outbreaks, including activity space data led to the detection of larger areas of higher spatial intensity and more focal points compared to residential location alone.

Interpretation: Geospatial analysis using activity space data (social gathering places as well as residence) may lead to improved understanding of areas of infectious disease transmission compared to using residential data alone.

Funding: This work was supported by funding from the National Institute of Allergy and Infectious Diseases R01AI097045, R01AI147336, and R01AI170204.

Keywords: Tuberculosis transmission; activity space; geographic heterogeneity; infectious disease control; outbreak; spatial analysis; whole genome sequencing.

PubMed Disclaimer

Figures

Figure 1.
Figure 1.
Histograms for distribution of number of activity space locations per participants for outbreak and genotypically ungrouped participants.
Figure 2.
Figure 2.
Posterior mean estimates and marginal distributions of the range and variance of the estimated spatial effect for each outbreak group, Gaborone, Botswana, 2012–2016
Figure 2.
Figure 2.
Posterior mean estimates and marginal distributions of the range and variance of the estimated spatial effect for each outbreak group, Gaborone, Botswana, 2012–2016
Figure 3.
Figure 3.
Posterior mean estimates of spatial random effect for each outbreak group (A-H) and controls (ungrouped participants), Gaborone, Botswana, 2012–2016. Values are shown on the internal linear predictor scale and represent the contribution of the spatial random effect on the response, after accounting for other fixed and random effects in the model. Departures from baseline (above or below zero) for outbreak groups measure group-specific spatial patterns that are not accounted for by the spatial distribution of activity spaces of controls. Darker colors correspond to increased spatial effect estimates. Values are displayed on the same color scale for all outbreak groups, though on a separate color scale for controls due to difference in sample size.
Figure 3.
Figure 3.
Posterior mean estimates of spatial random effect for each outbreak group (A-H) and controls (ungrouped participants), Gaborone, Botswana, 2012–2016. Values are shown on the internal linear predictor scale and represent the contribution of the spatial random effect on the response, after accounting for other fixed and random effects in the model. Departures from baseline (above or below zero) for outbreak groups measure group-specific spatial patterns that are not accounted for by the spatial distribution of activity spaces of controls. Darker colors correspond to increased spatial effect estimates. Values are displayed on the same color scale for all outbreak groups, though on a separate color scale for controls due to difference in sample size.
Figure 4.
Figure 4.
Predicted mean spatial intensity of activity spaces for participants in each outbreak group (A-H) and controls (ungrouped participants), Gaborone, Botswana, 2012–2016. Values are displayed on the response scale (obtained by exponentiating the linear predictor) and represent predicted numbers of activity spaces per unit area (approximately 0.25 × 0.25 km). Areas of increased intensity correspond to higher geographic concentration of activity spaces for participants in each group. Intensity values are displayed on the same color scale for all outbreak groups for ease of visual comparison, though on a separate color scale for controls due to difference in sample size.
Figure 4.
Figure 4.
Predicted mean spatial intensity of activity spaces for participants in each outbreak group (A-H) and controls (ungrouped participants), Gaborone, Botswana, 2012–2016. Values are displayed on the response scale (obtained by exponentiating the linear predictor) and represent predicted numbers of activity spaces per unit area (approximately 0.25 × 0.25 km). Areas of increased intensity correspond to higher geographic concentration of activity spaces for participants in each group. Intensity values are displayed on the same color scale for all outbreak groups for ease of visual comparison, though on a separate color scale for controls due to difference in sample size.
Figure 5:
Figure 5:
Exceedance maps for spatial effect greater than 0 (departure above baseline) with high probability (0.95) – full activity space and residential locations only, Gaborone, Botswana, 2012–2016.
Figure 5:
Figure 5:
Exceedance maps for spatial effect greater than 0 (departure above baseline) with high probability (0.95) – full activity space and residential locations only, Gaborone, Botswana, 2012–2016.
Figure 6.
Figure 6.
Exceedance maps for predicted spatial intensity to display ‘core areas’ by outbreak group based on full activity space and residential locations only, Gaborone, Botswana, 2012–2016
Figure 6.
Figure 6.
Exceedance maps for predicted spatial intensity to display ‘core areas’ by outbreak group based on full activity space and residential locations only, Gaborone, Botswana, 2012–2016
See this image and copyright information in PMC

References

    1. World Health Organization. Global Tuberculosis Report 2023. World Health Organization; 2023. Accessed March 3, 2024. https://www.who.int/publications-detail-redirect/9789240083851
    1. Vesga JF, Hallett TB, Reid MJA, et al. Assessing tuberculosis control priorities in high-burden settings: a modelling approach. Lancet Glob Health. 2019;7(5):e585–e595. doi:10.1016/S2214-109X(19)30037-3 - DOI - PubMed
    1. Auld SC, Shah NS, Cohen T, Martinson NA, Gandhi NR. Where is tuberculosis transmission happening? Insights from the literature, new tools to study transmission and implications for the elimination of tuberculosis. Respirol Carlton Vic. Published online June 5, 2018. doi:10.1111/resp.13333 - DOI - PMC - PubMed
    1. Shaweno D, Trauer JM, Doan TN, Denholm JT, McBryde ES. Geospatial clustering and modelling provide policy guidance to distribute funding for active TB case finding in Ethiopia. Epidemics. 2021;36:100470. doi:10.1016/j.epidem.2021.100470 - DOI - PubMed
    1. Zetola NM, Moonan PK, Click E, et al. Population-based geospatial and molecular epidemiologic study of tuberculosis transmission dynamics, Botswana, 2012–2016. Emerg Infect Dis. 2021;27(3):835–844. doi:10.3201/eid2703.203840 - DOI - PMC - PubMed

Publication types

LinkOut - more resources